Felix Geisendörfer (Datadog) - How to Win Frames and Influence Pointers

Go’s execution tracer (aka go tool trace) has suffered from high overhead since its inception in 2014. Historically this has forced potential users to worry about up to 20% of CPU overhead when turning it on. Due to this, it's mostly been used in test environments or tricky situations rather than gaining adoption as a continuous profiling signal in production.
This talk will explain the contributions we made to the Go 1.21 release that reduced this overhead to less than 1% for most applications using frame pointer unwinding. It will discuss the technical challenges we faced, the collaboration with the Go runtime team at Google, and how Go users can benefit from this work and improve the performance of their own applications.
Last but not least we’ll discuss other exciting improvements of the execution tracer in Go 1.22 and beyond.
Felix Geisendörfer is a Senior Staff Engineer at Datadog where he works on Continuous Profiling and contributes to the Go runtime. Before that he was working at Apple, co-founded Transloadit, contributed to node.js and inspired a generation of mad scientists to program flying robots with it. He is also a father, open source contributor, beach volleyball player, and giver of questionable advice on the internet.